Popular retellings of past bubbles always seem to give the impression that things were crazy back then, and that stocks were the only thing that people could think about as part of a “mass escape from reality”. While this idea has taken on bit of a life of its own to inform current discussions about the level of speculative excess in markets (or the lack thereof), I want to show in this blog post that ‘crazy’ is a relative term: People in the past were no more or less crazy than we collectively are today. Therefore, it’s essential to have an accurate view of cultural trends (both past and present) for us to maintain an objective view of complex systems like the stock market—that otherwise (blindly) aggregates information from the economic, financial, and cultural spheres.

Culture is what emerges from a great melting pot of memes

Richard Dawkins coined the idea of the ‘meme’ back in 1989, writing in The Selfish Gene that the meme is “a new kind of replicator” that emerged alongside the human brain as it evolved, providing a new, fertile ecosystem for thoughts and ideas to replicate in. A meme, Dawkins explains in the book, is a “unit of cultural transmission … a unit of imitation”, explaining that

Examples of memes are tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches. Just as genes propagate themselves in the gene pool by leaping from body to body via sperms or eggs, so memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense, can be called imitation. If a scientist hears, or reads about, a good idea, he passes it on to his colleagues and students. He mentions it in his articles and lectures. If the idea catches on, it can be said to propagate itself

Dawkins continues,

Imitation, in the broad sense, is how memes canreplicate. But … some memes are more successful in the meme-pool than others. … [and] particular examples of qualities that make for high survival value among memes … must be … longevity, fecundity, and copying-fidelity. … As in the case of genes, fecundity is much more important than [the] longevity of particular copies. If the meme is a scientific idea, its spread will depend on how acceptable it is to the population of individual scientists … if it is a popular tune, its spread through the meme pool may be gauged by the number of people whistling it in the streets. … Some memes, like some genes, achieve brilliant short-term success in spreading rapidly, but do not last long in the meme pool. Popular songs and stiletto heels are examples.

Ultimately, the ‘fitter’ a meme is, the more appealing it will be, and the more popular it becomes, the faster it will spread throughout the meme-pool that we call ‘culture’. At any given time, culture will be the emergent, collective average of all memes perpetuating through the collective human brain and which occupy their thoughts and habits. In some cases, these memes will be in the form of institutionalised knowledge (facts and paradigms, or ways of doing things), while others will manifest as faster-moving fads or trends. The resultant great melting pot of fast- and slow-moving memes and the churning of the mix over time is perhaps best represented by drawing an analogy with the Pace layering concept, where trends and fads represent the faster-moving fashion-layers, and where facts and paradigms are more fundamental and slower-moving.

In this modified Pace layering structure, memes exist on two axes, one measuring popularity (fads are more popular than reality) and the other measuring realism (reality is more realistic than fads). While popular but loosely anchored memes fizzle out quickly, themes with greater staying power can ‘drop’ into the layers below and increase their odds of being institutionalised as knowledge or enshrined as ways of doing things. 

A vague analogy can also be drawn here between this model and gene-based replicators like viruses, where fast-spreading viruses like SARS-CoV-2 can infect many people quickly, but whose biology means that it can never be more than a perennially circulating virus. Retroviruses like HIV, on the other hand, spread less quickly, but their biology allows them to occasionally be integrated into human genomes where they can be preserved as useless and decaying ‘genetic fossils’ for very long periods of time.

To describe the imitation strength of memes using words like ‘fecundity’ (as Dawkins does) however has the potential to be a bit misleading by taking the analogy between genes (or viruses) and memes too far. Instead, something like appeal (or Dawkins’ ‘acceptability’) might be more accurate, as concepts or ideas (in the form of memes) spread faster the more acceptable they are, and the stronger their appeal is, and the more people who happen to be receptive to the meme. 

The marketing and entertainment industries have of course been aware of this for a long time, as they specialise in generating and spreading memes, often by tapping into and by piggybacking on existing ones. For example, the blockbuster movie concept is built on the idea of near-universal appeal, as a production that is palatable to everyone (young and old, male and female, and across the world) is both de-risked and exposed to a much higher potential upside than any more specialised production (with a smaller audience and lower appeal) would be. Platform companies like Facebook, YouTube, or TikTok have taken this idea one step further by outsourcing the production of content to their users and allowing the users themselves to surface the content that is the most appealing to the largest number of people, while the platform contents itself with raking in the resultant advertising cash. 

The analogy between a Hollywood blockbuster and an outlier stock is also easy to draw, as both are generally acceptable and have a near-universal appeal. Because of this appeal, they also spread readily through word-of-mouth, except stocks are (typically) more likely to be recommended to you by your investor friends and the general industry than by your friends or neighbours. Blockbusters are also successful at the stock office for the same reason as outlier stocks are outliers: Because they are so popular, large numbers of people will be queuing up to be part of the show.

The stock market is Google Trends for stock-related memes

The analogy between blockbusters and outlier stocks and the ‘culture as a meme pool’-heuristic suggest that the stock market would be the ultimate platform for stock-related memes, where investors (individual or institutional) are using their money to cast short-term votes on individual stocks or companies. Despite temporary inefficiencies, where meme stocks like GameStop or Tesla can trade at steep premiums, in the long term—as Graham realised long ago—it is only the most stable and durable businesses will deliver long-term value for shareholders.

Accordingly, the stock market records the popularity of capital-market trends as they develop. In each era, the stock market has its own meme-driven investment trends and meta-trends (themes), and it’s by tracking these that the astute stock-market observer can effectively ‘take the pulse’ on the meme pool to intuit the direction of the economy-wide money flows. Sometimes, these flows are headed into the stock market as part of a thematic meta-trend (i.e. stocks are ‘hot’), while at other times, the flows are headed into other asset classes like housing (i.e. real estate is ‘hot’). During times when a lot of money is making its way into the stock market, mini-themes (trends) will result, where the money flows into the stocks from companies operating in a handful of industries or sectors, like renewables or electric vehicles or IT (software). Each theme will further see its own winners and losers, as some companies are better than the rest at capturing the imagination of investors—either by showing great fundamentals or succeeding at memetic marketing. (The very best stocks will have both great fundamentals and be generating attractive memes.)

Altogether, the ups and down of individual stocks will be recorded as stock-market price curves. While the random movement of these curves up and down are the product of random processes like investors balancing their portfolios or meeting client cash flow demands, the directed movement that develop in these curves (up or down) measure the changing popularity of the instruments in the investor meme pool. Because of this quality, where the trend or direction of a stock’s price changes reflect the aggregate outcome of investors’ capital-allocation decisions, each price curve, and all the market price curves in aggregate, mean that the stock market represents something like a ‘Google Trends’ for stocks. Here, the best-performing stocks are effectively the stocks with the most rapidly growing investor mindshare. These stocks represent the currently most successful and fittest finance-related memes circulating in the cultural meme pool. 

This model, where the stock market tracks the popularity of individual stocks as a proxy for the popularity of stock-related memes circulating in the meme tool also allows us to draw analogy to our modified Pace layering framework. Here, the fastest-changing and most volatile stocks would map onto the ‘fad’ layer (high in popularity and low on realism), showing a rapid price appreciation as the fad spreads through the meme pool and then dissipating, as quickly, as the fad burns itself out. The more realistic an investment thesis is, the slower-moving it is likely to be, as realism is a less attractive but more durable mimetic characteristic. 

As this suggests, our modified Pace layering model can act as a framework with two axes along which to analyse stocks: On the realistic axis, we’ll analyse stocks on the basis on fundamentals (where good fundamentals help seed memes), whereas on the popular axis, we’ll analyse stocks by assessing the attractiveness of their associated memes (where attractive memes help incentivise positive investment flows). Using this model, we’ll find that the price appreciation of some previously inexplicable outlier stocks start making a whole lot more sense: There are both good stocks and popular stocks, and while the ideal stock is both good and popular, there is much to be said for being able to analyse stocks both ways.

Bubbles result from meme pools being overweight stock-related memes

Sometimes the stock market itself becomes a meme. Indeed, when we consider the cultural backdrop across market cycles, we find that the meme pool is typically overweight stock-related memes during particularly frothy periods, and underweight during stock-market lulls. 

Regardless of the prevalence of stock-related memes, the underlying memetic diversity of the meme pool is however typically maintained. Instead, it’s the ratio of stock-related memes to other memes that goes up and down, and just because stocks are particularly popular during some periods and less so during other periods doesn’t mean that people stop enjoying gardening or talking about politics. 

I’m well aware that I’m mixing my metaphors pretty liberally here, but a useful mental model to use when visualising this dynamic (of the rise and fall of thematic memes) might be to draw analogy to how allele (gene variant) frequencies can fluctuate randomly over time in a biological population (gene pool), where an allele’s frequency varies by something approximating a random walk between generations.

The chart below shows the simulated frequency of a collection of alleles (gene variants) in the gene pool over time (I got the diagram from here, [link]). While all alleles start at a frequency of 50 %, random fluctuations between generations (depending on what carriers reproduce and which ones don’t) see the alleles either increasing or decreasing in frequency. The smaller the population is, the more volatile the fluctuations, and the larger the population, the less volatile the fluctuations.

Admittedly, this is a very imperfect way of visualising the frequency of memes in the meme pool—especially as there (by definition) is no autocatalytic element present in genetic drift (which would see the increased frequency of an allele increasing the frequency of that allele further), but hopefully you see what I’m getting at by referring to a modified version of the chart, below:

Here, if the blue line (for example) tracked the popularity of the stock-market meme in a culture (meme pool), it would be low at first (as only a few people would be interested in stocks and investing), but over time, due to random influences (albeit with the important autocatalytic component that’s missing here, where you’re excited by the stock market because all your friends are excited by the stock market), the frequency of the stock-market meme in the meme pool would increase. At the point where an above-critical number of people are host to this meme and are acting on it by investing in the stock market, the market would be experiencing a cyclical upswing. This would increase the prevalence of the stock-market meme further, as more people are ‘infected’ and want to join in on the fun. As this modified diagram also shows, even if one allele/meme is increasing/decreasing in popularity, there are typically multiple additional memes vying for attention, each experiencing its own cyclical up- and down-swings. Adding an autocatalytic element to this dynamic would just make the up- and down-swings more dramatic, as would happen if we were to increase the speed by which information and actions can be logged and processed (e.g. over the Internet, using free chat apps and zero-fees trading). In fact, the Financial Times has a great chart showing retail interest in different equity classes in this article here: [link].

Applying this model to the Pace layering framework above, suggests that fads would be memes with a wide (‘viral’) appeal but that score low on longevity, meaning that they very quickly burn their way through the meme pool before fizzling out to never be heard from again. (Aside from as part of future recollections, e.g. “Can you believe that Bill Gates did the ice-bucket challenge?”) Trends would be less volatile, but still fluctuating over time; never going ‘viral’ but also never sizzling out completely. Instead, some trends could become ‘fixed’ as more deep-rooted themes, while others would remain popular only among the select few. Something like the stock-market meme (where investing in stocks is seem as a sensible and popular thing to do) would be a periodically recurring trend that’s present at a high frequency during some periods and lingering at a low frequency during other periods.

Things weren’t really that crazy back then

Now, let’s return to the original topic of this post: How crazy were things, really, during past market-cycle peaks, and how does this compare to today’s craziness? 

The popular perception is that things were pretty crazy, and if we think back to the 1920s stock-market boom, the story typically goes that everyone was invested in the market. This view is perpetuated by famous stories of investors ‘selling their stock at the point where the shoe-shine boy was giving them stock tips’ because things, at that point, were believed to have gone too far. In reality, the public attitude was more muted, and if we turn to someone like John Kenneth Galbraith who chronicled the 1920s boom and bust in The Great Crash 1929, he says

There is probably more danger of overestimating rather than underestimating the popular interest in the market. The cliché that by 1929 everyone ‘was in the market’ is far from the literal truth. Then, as now [in 1954], to the great majority of workers, farmers, white-collar workers, indeed to the great majority of all Americans, the stock market was remote and a vaguely ominous thing. Then, as now, not many knew how one went about buying a security; the purchase of stocks on margin was in every respect as remote from life as the casino at Monte Carlo.

This view, Galbraith continues to say, is supported by a post-crash “Senate committee investigating the securities market to ascertain the number of people who were involved in securities speculation in 1929”, finding that “only one and a half million people, out of a population of approximately 120 million … had an active association of any sort with the stock market” and that “only 600,000 of these accounts … were for margin trading, as compared with roughly 950,000 in which trading was for cash”. 

Even if we should take Galbraith (who wrote his account of the 1929 crash in 1954) with a bit of salt, his numbers suggest that only about 1 % of Americans were carrying the stock-market meme and acting on it. Regulatory changes (including the introduction of the 401(k) plan in the 1980s) have of course changed this dynamic somewhat, as more investors are vaguely aware of the stock market today than they would have been in the past (Pew Research numbers from 2020 suggests that over 50 % of US households were invested). That’s pretty crazy (given how risky stocks are), and it shows that the stock-market meme is pretty prevalent (and probably with a higher floor today than in the past, because of the need for so many people to grow their retirement savings). Indeed, the IPO market has been strong for some time now, and the SPAC craze was running hot for a while. In addition, we have been seeing retail investors using the stock market as a way to ‘get back’ at Wall Street. All of these are signs that the market cycle is peaking.

One counter-meme that is symptomatic of a high prevalence of the stock-market meme is the need to justify the market’s good performance. A popular version of this counter-meme is to point to the bears and say that ‘If these people are saying that we are in a bubble, then we cannot be in one, as being in a bubble means that everyone is a believer’. This counter-meme is, as I hope to have shown with the quotes above, a direct symptom of the popular misconception that things need to be as crazy as the tales of bubbles past to signal that we are in one. This, of course, isn’t true, and popular retellings typically compress a decade’s worth of stock-market history into a sentence or two, like: ‘A booming economy led to growing investor enthusiasm, which eventually turned into euphoria. Investors got ahead of themselves and traded too much on margin, and when the economy started dipping, the stock market crashed.’ The choice of the word ‘euphoria’ is particularly damaging to perception, as it brings to mind investors cheering the market on with champagne, something that we don’t really see around ourselves today (or I’ve been spending time with the wrong people…). 

Instead, a more accurate definition of ‘euphoria’ is the growing feeling that the market cannot go down. This feeling typically develops after years of reliable stock market returns, where the market seems to go only up and up, and more reliably so for each year that passes. Over time, the idea that the market can go down fades out of the popular consciousness as people forget. Looking at the market, such short-sightedness can be understandable: When the market is going up, it’s hard to realise how volatile stocks can be. Over time, reliable returns breed complacency, and complacency encourages investors to take on more and more risk because they don’t think they’ll end up paying for it. This is euphoria. 

One sign that euphoria has taken root is the strong conviction that it’s advisable to ‘buy the dip’. In the short term, this is sensible advice, as the further along the market gets to its peak, the more common rocky patches become. These drawdowns (typically of 20 – 40 % at a time) quickly revert as there are plenty of investors being willing to ‘buy the dip’ and to prop the market up before propelling it to greater heights. At some point, this reserve buying-power however gets increasingly exhausted, and at the point where investors are tapped out and exhausted with the stock-market meme, they will start looking to invest their dollars in other assets. That way, when a ‘correction’ comes, there are too few people to prop the market up, and it starts dipping—and then declining further. And the further it falls, the less likely people are to put their dollars at risk, as other asset classes start looking more attractive (and other memes start increasing in prevalence at the expense of the stock-market meme). 

Past experiences of bubbles popping also tell a story of extremely anti-climatic market peaks: The peak is typically not accompanied by much fanfare, and often pass quietly, without much comment even by the bubble-callers themselves. For example, Galbraith writes in The Great Crash that

On 3 September, by common consent, the great bull market of the nineteen-twenties came to an end. Economics, as always, vouchsafes us few dramatic turning points. Its events are invariably fuzzy or even indeterminate. One some days that followed – a few only – some averages were actually higher. However, never again did the market manifest its old confidence. The later peaks were not peaks but brief interruptions of a downward trend. ¶ On 4 September, the tone of the market was still good.

It was only later (admittedly a day or so), when stocks dropped more dramatically, that people started to get worried. 

John Cassidy in his (surprisingly enjoyable) 2004 book on the dotcom boom and bust, dot.con, tells a very similar story, where, even as the market was getting increasingly worried in early April 2000 (the NASDAQ peaked in late March), there were still differences of opinion and plenty of people who were willing to ‘buy the dip’.

James Cramer claimed the worst was over … Business Week rejoined the stock market boosters with a cover that asked: ‘WALL STREET: IS THE PARTY OVER?’ The answer, also emblazoned on the front page, was unequivocal: “High-tech stocks are undergoing a much-needed correction. But relax, the overall market probably won’t tank. What we’re seeing looks more like a healthy flight to quality.

Justifications and excuses accompany stocks on the way up

Because not everyone will be ‘infected’ with the stock-market meme, even at the market peak (as even the most popular meme will fail to ‘infect’ every single brain), there will—at both early and late stages of the bubble—be plenty of people willing to point out the excesses of other market participants. Towards the end of a market cycle, it’s sometimes even the people with the most to lose who will make the most candid comments. 

What sets today apart from other market peaks is that there was a general feeling of insecurity in the market before it peaked in both 1929 and 2000, that was making even long-time bulls uneasy. Other parts of the historic narrative however fits well with what we’re seeing today, where even market cheerleaders like ARK Invest’s Cathie Wood are warning of the potential for near-term corrections:

Rejecting talk of an equity market bubble led by many of their high profile holdings, Wood said she expects innovative companies will grow into their high equity valuations as demand for their products and services expands exponentially over the next decade. Aside from Tesla, Wood reckons the broad adoption of digital wallets will benefit the payments groups Square and PayPal, for example. Wood said a stock market correction will provide Ark with an opportunity to buy more “high conviction names” and companies it believes will be “winner-take-all” in a fast-expanding new industry. “A correction is a great time to determine what are our high conviction names,” she said.

Just for fun, let us compare what Wood is saying to statements made by Morgan Stanley’s Mary Meeker towards the end of the dotcom bubble (in May 1999):

In a research report, Morgan Stanley’s Mary Meeker said she expects “more weakness” from Internet stocks, which will likely make investors more cautious about the sector.

“It isn’t uncommon this time of year for prices of technology stocks to contract when catalysts are insufficient,” Meeker wrote.

Meeker also noted she wouldn’t be surprised if Morgan Stanley’s Internet index — already down 33 percent from its April high — fell another 20 percent.

Superstar sell-side analysts also admitted during the dotcom boom that some growth stocks were “extremely expensive” as early as in 1998, but kept finding various ways to justify their increasingly ambitious price targets. Merrill Lynch’s Henry Blodget, for example, wrote that

The overall Internet stock phenomenon may well be a ‘bubble’, but at least in one respect it is very different from other bubbles: there are great fundamental reasons to own these stocks … the companies underneath these stocks are (1) growing amazingly quickly, and (2) threatening the status quo in multiple sectors of the economy.

Here’s Blodget again (from dot.con, like the previous quote):

When Yahoo! went public [in 1996], it looked like the biggest joke in history … Investors the world over (understandably) crowed about manias and insanity, but [Yahoo!] was actually trading at an insanely cheap 10X Q4 1998 annualised earnings. Investors who failed to ask themselves two questions—(1) how big the company could actually be, and (2) how fast it could get there—missed the boat. With these types of investments, we would also argue that the real ‘risk’ is not losing some money—it is missing a much bigger upside.

While written in 1998, these words are almost verbatim from multiple conversations that I’ve personally had with growth investors over that past couple of years. ‘Valuations’, they say, ‘are justified’, one way or another, whether it’s through ‘embracing uncertainty’ or ‘imagining the upside’ or even making value-based appeals like ‘corporate earnings have never looked so good’. Of course, this type of behaviour (especially when coupled with strategies like running your winners, however high they soar into the stratosphere) is perfectly justified when lots of stocks are going up and even dart-throwing chimpanzees can pick a market-beating portfolio of stocks.

What this lesson from history shows is that just because some people are calling the market expensive, or even as being in bona fide bubble territory, this is not, by itself, evidence that we are notin a bubble. In fact, bubbles are easy to spot—even by the people who are the most invested in them. Talk about bubbles are therefore not a symptom of a healthy market. Instead, the bubble meme (another counter-meme) grows more prevalent the more prevalent the market-meme itself becomes.  

Below are two charts of the NASDAQ Composite, on which we can map the statements above.

Something that’s not tracked in any of these charts are the drivers of each market cycle, which dependend on unique, era-dependent factors. (The stock market, like Google Trends, tracks only the symptom of memes spreading in the population, and not their exact nature of what precipitated their spread.) Notably, in March 2000, the market was already skittish following a growing number of penetrating analyses of cash-losing Internet stocks; analyses that showed that the times were already turning, and that the stock-related memes were burning out. Therefore, when the market started dipping, there were fewer people infected with the market meme to be around and actively buying the dip: Because the meme was losing its appeal, stocks just weren’t as cool anymore as they’d been before. Instead, more money started rotating into more tangible assets like real estate (potentially seeding the housing bubble), but also into tangible assets (which led to higher inflation, which spooked the Fed). 

It doesn’t feel like we’re quite at that point yet, but then again, if we’re in a bubble today, it is a more grown-up bubble: We haven’t seen quite the level of excess that we saw during the dotcom era and a lot of the existing growth stocks have been around for much longer than the year-old companies that were IPOing during the height of the dotcom boom. So maybe things actually were crazier back then, after all? But then again, the dotcom boom saw AOL become the most valuable entertainment company in the world (which allowed them to acquire Time-Warner), and there were also loss-making delivery operations that used student bicyclists to deliver items of convenience to residents in NYC. Both of these fateful business ideas are 20-year old premonitions of some of the excesses that we’re seeing in markets today. 

Ultimately, I guess we’ll have to wait a while longer for history to make its verdict on how today’s craziness is stacking up.